Nakh-Daghestanian linguistic family ; Celtic & Afro-Asiatic langua

I thought some of you might be interested in this commentary article from
the current issue of Science (forwarded by Randy Foote). Comments? -- Marc
Verhaegen



LINGUISTICS:
Peering Into the Past, With Words
Bernice Wuethrich
Prehistorians typically rely on stones, bones, and DNA to piece together the
past, but linguists argue that words preserve history too. Two new studies,
both based on endangered languages, offer new insights into the identity of
mysterious ancient peoples, from the first farmers to early inhabitants of
the British Isles.

Archaeologists have long known that some 10,000 years ago, ancient people in
Mesopotamia discovered farming, raising sheep, cattle, wheat, and barley.
And
researchers knew that by 8000 years ago agriculture had spread north to the
Caucasus Mountains. But they had little inkling of whether traces of this
first farming culture lived on in any particular culture today. People have
migrated extensively through the region over the millennia, and there's no
continuous archaeological record of any single culture. Linguistically, most
languages in the region and in the Fertile Crescent itself are relatively
recent arrivals from elsewhere.

Now, however, linguist Johanna Nichols of the University of California,
Berkeley, has used language to connect modern people of the Caucasus region
to the ancient farmers of the Fertile Crescent. She analyzed the
Nakh-Daghestanian linguistic family, which today includes Chechen, Ingush,
and Batsbi on the Nakh side and some 24 languages on the Daghestanian side;
all are spoken in parts of Russia (such as Chechnya), Georgia, and
Azerbaijan.

Nichols had previously established the family tree of Nakh-Daghestanian by
analyzing similarities in the related languages much the way biologists
create a phylogeny of species. She found that three languages converge at
the
very base of the tree. Today, speakers of all three live side by side in the
southeastern foothills of the Caucasus Mountains, suggesting that this was
the homeland of the ancestral language--on the very fringes of the Fertile
Crescent. To get a rough estimate of when the language arose, Nichols used a
linguistic method that assumes a semiregular rate of vocabulary loss per
1000
years, and she dated the ancestral language to about 8000 years ago.

Nichols also found that the ancestral language contains a host of words for
farming. The Chechen words muq (barley), stu (bull), and tkha (wool), for
example, all have closely related forms in the earliest branches of
Daghestanian, as do words for pear, apple, dairy product, and oxen yoke--all
elements of the farming package developed in the Fertile Crescent. Thus
location, time, and vocabulary all suggest that the farmers of the region
were proto-Nakh-Daghestanians. "The Nakh-Daghestanian languages are the
closest thing we have to a direct continuation of the cultural and
linguistic
community that gave rise to Western civilization," Nichols says.

Population geneticist Henry Harpending of the University of Utah, Salt Lake
City, has just begun the job of unraveling the genetic ancestry of
Daghestanian speakers and is impressed with Nichols's work. "For years I
wished linguists would get in the game. Nichols sure is."

Nichols is now reconstructing the ancestral language, hoping for more clues
to the culture of these early farmers. But she has to work fast, for the
three Nakh languages are vanishing. Although there are still about 900,000
Chechen speakers left, the other two tongues have fewer speakers, and all
three are being eroded by war, economic chaos, and Russian educational
practices, Nichols says.



More than 3200 kilometers away, another linguist is mining Celtic
languages--which are also all considered endangered--for clues to the
inhabitants of the early British Isles. Artifacts show that the islands were
occupied long before Celts from the European continent made landfall about
700 B.C. But mysteries remain as to their identity.

So Orin Gensler of the Max Planck Institute for Evolutionary Anthropology in
Leipzig, Germany, analyzed Celtic languages, including Irish Gaelic,
Scottish
Gaelic, Welsh, and Breton. Once prevalent throughout Europe, these languages
are now spoken only in the British Isles and Brittany in France. Linguists
have noted surprising grammatical differences between Celtic languages and
related languages such as French, while at the same time seeing striking
resemblances between Celtic and Afro-Asiatic languages spoken for millennia
across a swath of coastal Northern Africa and the Near East.

In a forthcoming monograph, Gensler studied 20 grammatical features found in
both Celtic and Afro-Asiatic languages. He sought these linguistic traits in
85 unrelated languages from around the world, reasoning that if the features
were widespread, their appearance in both Celtic and Afro-Asiatic languages
might be mere coincidence. But if the shared features are rare, coincidence
is unlikely. Overall, Gensler found that about half the shared features are
rare elsewhere. "I think the case against coincidence is about as good as it
could be," he says.

And a closer look at a number of features, including word order, offers a
provocative theory for just how the Celtic islanders acquired these
linguistic traits. In Gaelic and Welsh--and many Afro-Asiatic languages--the
standard sentence structure is verb-subject-object. But Celtic languages
spoken in Continental Europe in antiquity have the verb in the final or
middle position. The best explanation for the shift to verb-initial order,
says Gensler, is that when Celtic speakers made landfall on the British
Isles, Afro-Asiatic speakers were already there. As these people learned
Celtic, they perpetuated aspects of their own grammar into the new language.

Although others are interested in Gensler's idea, so far "there is no
significant northwest African genetic signature ... in Celtic populations,"
says Peter Underhill, a molecular geneticist at Stanford University in
California. But in this instance, he adds, the linguists may be ahead of the
geneticists, for researchers need more genetic markers before they can
confirm or refute Gensler's idea.